Operation of the S128240[C/D] in grey-scale mode
The S128240[C/D] display uses the Sitronix ST7529 controller.
This controller is capable of 5-bit-grey-scale operation.
E.g. instead of just having a pixel on/off, the shade of the pixel could be set to 30 grey values in addition to the basic 'on/off'.
I'm wondering whether the S128240 display and nematic is designed to work with this feature of the controller?
I'm asking this as when I tested the feature with the standard init_LCD(), I noted two anomalies.
1) The display was severely 'streaked' in columns which also contained 'grey'
2a) When ~30%of the display is set to grey, the charge pump can't keep-up
2b) When ~10% of the display is set to grey, and there is significant black on the display, the charge pump can't keep up.
TEST1 - PASS
Clear screen (write 0x00 to each pixel)
Print a line of text in 100% intensity at the top of the screen I.e. writing 0xff to each 'on' pixel
Paint a 32 x 240 graphic in 100% intensity (black) at the top of the screen I.e. writing 0xff to 'on' each pixel
RESULTS: The graphic and text were clear, distinct with uniform pigmentation.
TEST2 - FAIL
Repeat test 1, keeping 100% intensity of the text, but changing the graphic to 75% intensity I.e. writing 0xc0 to each 'on' pixel
RESULTS: The graphic was banded, and faint, text was streaked where the graphic was most intense.
This test was repeated at several other intensity settings, all with the same result.
TEST3 - PASS
Clear screen (write 0x00 to each pixel)
Print a line of text in 100% intensity at the top of the screen I.e. writing 0xff to each 'on' pixel
Paint three instances of a 64 x 64 solid block in 100% black across the center-line of the screen (12 pixels between each block laterally)
RESULTS: The graphic and text were clear, distinct with uniform pigmentation.
TEST4 - FAIL
Repeat test3, keeping 100% intensity for the text, but changing the blocks to 75% intensity
RESULTS: All of the display faded-out. The text was streaked (fading out in the same columns as the blocks).
Note, the text used was from a 5 wide x 12 high font.
To summarize
1) Can the S128240 be operated in 'grey-scale'
2) Are there any optimizations to the LCD_init required for this?
Hi Martin,
Out of curiosity, did you ever trying by-passing the charge pump and using greyscale? The S128240D should be able to handle greyscale, however the charge pump is always going to be the weak link is a display of this size, probably not so much of an issue on the S128240C. There aren't any optimization tricks to the intialization for greyscale that we are aware of.
Regards,
Trevor
Hi Trevor,
There may be two separate issues at work.
1) As you stated, the charge pump is struggling with grey on the large format display
2) However, there also appears to be another issue (possibly with the voltage follower / drivers / nematic)
DETAIL
test2 behaved in the same way (streaking) for the C/D displays and with / without internal bias generation
For the 'D' display the internal change pump was generating Vlcd of ~16V and V0in of 15.11V.
15.11 is the voltage I'd programmed in.
test4 behaved differently for the 'C' and 'D' with/without internal bias.
All combinations exhibited the streaking problem.
However the fade-out was only exhibited on the 'D' display with internal charge-pump.
This was corroborated by the 'D' Vlcd being dragged down to ~14.3V.
I'd half expect there to be issues with the charge-pump on a large format display, so the test4 results are not surprising.
However, would you be able to confirm the test2 results? I.e. the non-charge-pump issue (the image from the S128240C contrast thread works well with an 'on' pixel value of 0xc0 instead of 0xff).
Regards
Martin
